Hayden Field

OpenAI hired Dylan Scandinaro, an AGI safety executive from competitor Anthropic, as its new head of preparedness.

“Author of "OpenAI Poached Its New Safety Executive From Anthropic" in The Verge”

OpenAI is finally allowing current and former employees to donate their equity to charity after an 18-month delay, potentially enabling six-figure equity holders to donate millions.

“Author of "OpenAI is Finally Letting Employees Donate Their Equity to Charity" in The Verge”

OpenAI CEO Sam Altman says there are no current plans to introduce ads in ChatGPT Pulse, but he has not ruled out adding relevant advertising in the future similar to Instagram's model.

“Author of "Sam Altman Says There Are No Current Plans for Ads Within ChatGPT Pulse — But He's Not Ruling It Out" in The Verge”

California Governor Newsom signed SB 53, a landmark frontier AI transparency law requiring large AI companies to publicly disclose safety frameworks and report critical incidents, with whistleblower protections and annual regulatory updates.

“Author of "California Governor Newsom Signs SB 53, First Frontier AI Transparency Law" in The Verge”

Apple downplayed AI at its iPhone 17 event despite heavy AI emphasis last year, focusing instead on hardware improvements and background AI capabilities while lagging competitors in agentic AI development.

“Author of "Apple Just Ghosted AI At Its Biggest Event"”

Anthropic will begin training its AI models on user chat transcripts and coding sessions by default unless users opt out by September 28, 2025, while extending data retention to five years.

“Author of "Anthropic Will Start Training Its AI Models on Chat Transcripts" in The Verge”

Anthropic will begin training its AI models on user chat transcripts and coding sessions by default, with an opt-out option and a five-year data retention policy, requiring all users to make a decision by September 28th.

“Author of "Anthropic's New Update Puts You in Control" in Anthropic”

Anthropic released a memory function for Claude that allows users to search and reference past conversations, though it only activates when explicitly requested, unlike OpenAI's persistent memory approach.

“Author of "Anthropic's Claude Chatbot Can Now Remember Your Past Conversations" in The Verge”